For those still confused about the 2-pt call

[Kevinicus]

I hope this helps you understand.

 

[xwalker]

I hope this helps you understand.

The sample size is too small.

They need many years of teams actually implementing that strategy to get sufficient data.

 

[xwalker]

I hope this helps you understand.

Schoonmaker had a 100% catch rate vs Tampa Bay.

 

[Cowboys5217]

It was a brain dead stupid call, and it'll be one of the calls that gets Schotty fired at the end of the year.

 

[Kevinicus]

The sample size is too small.

They need many years of teams actually implementing that strategy to get sufficient data.

The math of the probabilities still works regardless of sample size. But yes, there should be more data. Unfortunately, people are stuck in old ideas they don't want to look at the reality of that probabilities part, and just want to stick with what they know.

 

[Kevinicus]

It was a brain dead stupid call, and it'll be one of the calls that gets Schotty fired at the end of the year.

Was the video confusing?

 

[Tabascocat]

Analytically, it was the correct call. But, situationally it was dumb at the time. The offense wasn’t very good that game and two EP’s was much easier than converting one two-point conversion out of two. It is playing for OT either way, no reason to go for two in that instance.

 

[JoeyBoy718]

Good post.

The sample size is too small.

They need many years of teams actually implementing that strategy to get sufficient data.

The sample size isn’t really small. It’s math.

For the sake of argument, let’s say you have a 100% chance of making an extra point and a 50% of winning in overtime.

These are reasonable assumptions. If two teams play to overtime, it’s fair to say it’s anyone’s game. Of course momentum matters, but there’s also chance with the coin flip.

So you have 3 strategies:

Strategy 1:
Kick XP, Kick XP, Go to overtime.
That’s 100% x 100% x 50% = 50%

Strategy 2:
Kick XP, Go for 2
That’s the likelihood of converting the 2

Strategy 3:
Go for 2, Kick XP or Go for 2
That’s the likelihood of converting the 2, but you get two chances

Strategy 3 is clearly better than Strategy 2 because you have an extra chance.

Let’s pretend that you have a 50% chance of converting the 2. That would make Strategy 1 and 2 the same.

So what are the odds of winning with Strategy 3?

It’s the same odds of flipping a coin twice and getting heads at least once.

You have 4 scenarios when flipping a coin:
1. Heads and heads
2. Heads and tails
3. Tails and heads
4. Tails and tails

3 of the 4 have at least one heads. So the probably of winning the game in Scenario 3 is 75%.

In fact, even if you only had a 29.3% chance of converting the 2, your chances still become 50% when you get a second attempt.

And, like the video said, most teams average between 38-48% on conversions.

 

[big dog cowboy]

This is EXACTLY what I said last week.

Option 3 increases your odds of winning. Isn't winning the goal of playing?

I was right then and am still right now.

 

[big dog cowboy]

It was a brain dead stupid call, and it'll be one of the calls that gets Schotty fired at the end of the year.

Did you watch the entire video and understand?

 

[fivetwos]

You can’t even compare one to the next on this.

The matchups and situations are entirely different. Regardless of what other teams have done in the past, that tells you nothing about your immediate situation.

Just ridiculous.

 

[JoeyBoy718]

Your chances are significantly higher in Scenario 3. Even more so now that they’ve moved the extra point back.

 

[Captain43Crash]

Who needs a human coach, just have a Ai robot!

I completely disagree with automatically going for 2 in this situation. It is WAY smarter for the head coach, to make the decision on going for 2, based on how confident your offense is and more importantly, how well your offense has been moving the ball vs the particular defense. Shottenheiner had NO answers for TB’s blitzing. On the long drive to get within 8 pointsTB played a more conservative prevent defense. They made it easier for Dak to score. Dak had also been confused most of the game.
By watching our offense struggle almost the whole game vs TB’s defense, I had very little confidence in us converting the 2 point conversion.
What happened? The defense that had been kicking our butt for 80% of the game stopped us on the 2 point conversion.

That should not surprise anyone, including the analytics robots!

Last week the Giants were down by 14 to the Commanders. The Giants were controlling the game late. They kicked 2 extra points to tie the game. Analytics may have said go for 2, but Harbough probably felt why take any unnecessary risk. We have the momentum we can beat them in OT.

They would have won in OT, but Skateboo fumbled right before they had a chance to kick a short game winning FG.

 

[xwalker]

Good post.

The sample size isn’t really small. It’s math.

For the sake of argument, let’s say you have a 100% chance of making an extra point and a 50% of winning in overtime.

These are reasonable assumptions. If two teams play to overtime, it’s fair to say it’s anyone’s game. Of course momentum matters, but there’s also chance with the coin flip.

So you have 3 strategies:

Strategy 1:
Kick XP, Kick XP, Go to overtime.
That’s 100% x 100% x 50% = 50%

Strategy 2:
Kick XP, Go for 2
That’s the likelihood of converting the 2

Strategy 3:
Go for 2, Kick XP or Go for 2
That’s the likelihood of converting the 2, but you get two chances

Strategy 3 is clearly better than Strategy 2 because you have an extra chance.

Let’s pretend that you have a 50% chance of converting the 2. That would make Strategy 1 and 2 the same.

So what are the odds of winning with Strategy 3?

It’s the same odds of flipping a coin twice and getting heads at least once.

You have 4 scenarios when flipping a coin:
1. Heads and heads
2. Heads and tails
3. Tails and heads
4. Tails and tails

3 of the 4 have at least one heads. So the probably of winning the game in Scenario 3 is 75%.

In fact, even if you only had a 29.3% chance of converting the 2, your chances still become 50% when you get a second attempt.

And, like the video said, most teams average between 38-48% on conversions.

I have a graduate degree in mathematics.

Your explanation is not correct.

 

[JoeyBoy718]

Actually, my math was a bit off. You have a 75% of converting on one of the attempts, but if you fail on the first, you go to overtime.

Using last year’s season averages:

XP conversion: 96%
2P conversion: 41%

And using 50% chance of winning in overtime (this obviously changes in a real game, but it’s the only baseline we can use).

Strategy 1:
XP * XP * OT = 0.96 * 0.96 * 0.5 = 46%

Strategy 2:
XP * 2P = 0.96 * 0.41 = 39%

Strategy 3:
2P * (XP or 2P). If you convert on the first 2P, you kick to win. If you convert on the second 2P, you go to overtime.
The formula is more complicated, but it’s 52%. Still higher than the others, but not as significant.

 

[JoeyBoy718]

I have a graduate degree in mathematics.

Your explanation is not correct.

You’re right. I didn’t factor in the fact that you still need to go to overtime if you fail to convert the first 2P attempt. I corrected it below. It’s 52%

 

[RonWashington]

I still think 24-16 you kick your extra point go down 7 which doesn’t force you later on to go for 2 . Aubrey has been automatic with XP I trust him with those . I think in the context of that game how is your offense performing and for an entire 4 quarters the offense struggled . First things first go down 7.

 

[Kevinicus]

Who needs a human coach, just have a Ai robot!

I completely disagree with automatically going for 2 in this situation. It is WAY smarter for the head coach, to make the decision on going for 2, based on how confident your offense is and more importantly, how well your offense has been moving the ball vs the particular defense. Shottenheiner had NO answers for TB’s blitzing. On the long drive to get within 8 pointsTB played a more conservative prevent defense. They made it easier for Dak to score. Dak had also been confused most of the game.
By watching our offense struggle almost the whole game vs TB’s defense, I had very little confidence in us converting the 2 point conversion.
What happened? The defense that had been kicking our butt for 80% of the game stopped us on the 2 point conversion.

That should not surprise anyone, including the analytics robots!

Last week the Giants were down by 14 to the Commanders. The Giants were controlling the game late. They kicked 2 extra points to tie the game. Analytics may have said go for 2, but Harbough probably felt why take any unnecessary risk. We have the momentum we can beat them in OT.

They would have won in OT, but Skateboo fumbled right before they had a chance to kick a short game winning FG.

So, you'd have confidence in the offense in OT? Not after marching 99.5 yards?

 

[Kwyn]

The sample size is too small.

They need many years of teams actually implementing that strategy to get sufficient data.

Not to be flippant, but the math and probabilities of each scenario don’t need a larger sample size because it’s math.

If a team converts over 38% of 2-point attempts, going for it after the first score mathematically guarantees better odds.

There are thousands of NFL data points proving the league averages 48%.

The main pushback is that momentum and coaches feel is more important than the analytics

The problem with that comes down to two things.

First, momentum is largely based on hindsight. It’s more a feeling about the play that just happened than it is about what happens next. It’s part of our human brain to assign meaning to things.

Second, even if momentum is real, going for it after the first score actually builds it. If you make it, the team immediately feels they are going to win.

Miss it and the team stil has the sense that they have a good chance to win because they get a second crack at it.

 

[diamonddelts]

The math of the probabilities still works regardless of sample size. But yes, there should be more data. Unfortunately, people are stuck in old ideas they don't want to look at the reality of that probabilities part, and just want to stick with what they know.

Analytics don't win football games. Men do.